Human-aware navigation system for assistive wheelchair

• Main Author: Chan, Maximilian Jun Feng
• Other Authors: Ang Wei Tech
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Mechanical engineering::Assistive technology
• Online Access: https://hdl.handle.net/10356/168316
• Institution: Nanyang Technological University

Assistive wheelchair is a type of wheelchair which assists individuals who are physically challenged in coping with their illness, injury or disability. Assistive wheelchairs are becoming prevalent in our society today, especially amongst the healthcare industry. There is an increasing need for an autonomous wheelchair following solution due to various reasons, mainly stemming from the issue of difficulty and safety. This project explores the need for wheelchairs to follow humans around. Traditionally, human following for wheelchair detects and tracks a target that would be walking ahead of the wheelchair itself. However, this project focuses on a side following approach where the wheelchair would detect and track a target that would be walking side by side with it. A side following solution would be difficult to implement due to various reasons such as the limitations of the detection and tracking algorithm, the position of the camera or sensors, and even the problem of self-occlusion of the target user itself. Specifically, this focuses on a specific use case: A generic follow-me solution in a hospital setting. This proves to be an important and focal aspect that should be looked at as a hospital setting revolves around multiple tight and constrained spaces, within a crowded environment with multiple people walking about, rendering a wheelchair following solution to be tricky and hard to handle. Thus, this project aims to explore solutions to enable side following solutions within a hospital environment. There are many various technologies available to detect and track humans. This project explores some of the technologies available in the market, and do comparisons before selecting the two best technologies suited for our solution and our use case. The two technologies selected will be prepared and set up, before experiments are conducted on both technologies, to determine the technology that best targets our use case. Thereafter, the two technologies will be integrated with the wheelchair and experiments will be further conducted to test how well the whole system would fare to detect and track targets through a side following approach.